This is just a section of the larger revised Title VI Legal Manual. Please click here to see the complete revised Manual.
Title VI Legal Manual
SECTION VI: PROVING DISCRIMINATION – INTENTIONAL DISCRIMINATION
Table of Contents
Title VI prohibits discrimination based on “race, color, or national origin …under any program or activity receiving Federal financial assistance.” 42 U.S.C. § 2000d. The purpose of Title VI is simple: to ensure that public funds are not spent in a way that encourages, subsidizes, or results in discrimination on these bases. Toward that end, Title VI bars intentional discrimination. See Guardians Ass’n v. Civil Serv. Comm’n, 463 U.S. 582, 607–08 (1983); Alexander v. Choate, 469 U.S. 287, 292–93 (1985). A Title VI discriminatory intent claim alleges that a recipient intentionally treated persons differently or otherwise knowingly caused them harm because of their race, color, or national origin. Agency regulations implementing Title VI also prohibit intentional discrimination based on race, color, or national origin, covering any disposition, service, financial aid, or other benefits provided under the recipient’s program, the determination of the site or location of facilities, or other aspects of program operations. See, e.g., 28 C.F.R. § 42.104(b) (Department of Justice regulations).
Private parties seeking judicial enforcement of Title VI’s nondiscrimination protections must prove intentional discrimination. Alexander v. Sandoval, 532 U.S. 275, 280–81 (2001). Private parties may also file administrative complaints with federal agencies alleging that a recipient of the agency’s federal financial assistance has engaged in intentional discrimination; the federal agency providing the assistance may investigate these complaints.
This section provides an overview of the types of evidence necessary to prove intentional discrimination under Title VI. Much of the discussion in this section relies on judicial precedent developed in private plaintiffs’ intent claims for damages, and therefore focuses on standards applied in that context. Those standards may not always apply to agency investigations, which often follow a non-adversarial model in which the agency collects all relevant evidence and then determines whether the evidence establishes discrimination. Under this model, agencies do not “shift the evidentiary burdens” between complainant and recipient when making findings. The burden-shifting framework may nevertheless serve as a useful paradigm for organizing and analyzing the evidence.
AGENCY PRACTICE TIP
Investigating agencies can look to case law for guidance on proving intentional discrimination, but are not bound by case law concerning burden shifting between plaintiff and defendant (that is, as between a complainant and a recipient). An agency need not use the same sequential process as courts, where a plaintiff first offers prima facie evidence and the defendant then offers rebuttal evidence. Rather, an agency has discretion to gather and evaluate all relevant evidence as part of its initial investigation, or may choose to make a preliminary prima facie finding then require recipients to articulate defenses.
B. Proving Intentional Discrimination
Courts have developed a number of analytical frameworks for assessing intent claims. The elements of a Title VI intent claim derive from and are similar to the analysis of cases decided under the Fourteenth Amendment’s Equal Protection Clause  and Title VII of the Civil Rights Act of 1964, as amended. Because the Title VI statutory prohibition on discrimination is based on the Equal Protection Clause, the constitutional analysis of intentional discrimination should be applied under Title VI. See Grutter v. Bollinger, 539 U.S. 306, 343–44 (2003) (citing Regents of Univ. of Cal. v. Bakke, 438 U.S. 265, 287 (1978) (opinion of Powell, J.) (“Title VI . . . proscribe[s] only those racial classifications that would violate the Equal Protection Clause or the Fifth Amendment.”).
Generally, intentional discrimination occurs when the recipient acted, at least in part, because of the actual or perceived race, color, or national origin of the alleged victims of discriminatory treatment. Doe ex rel. Doe v. Lower Merion Sch. Dist., 665 F.3d 524, 548 (3d Cir. 2011). While discriminatory intent need not be the only motive, a violation occurs when the evidence shows that the entity adopted a policy at issue “‘because of,’ not merely ‘in spite of,’ its adverse effects upon an identifiable group.” Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 279 (1979). Some assume that the intentional use of race should be carefully scrutinized only when the intent is to harm a group or an individual defined by race, color, or national origin. That is not true: the Supreme Court in City of Richmond v. J.A. Croson Co., 488 U.S. 469, 493 (1989), and Adarand Constructors, Inc., v. Pena, 515 U.S. 200, 226 (1995), established that any intentional use of race, whether for malicious or benign motives, is subject to the most careful judicial scrutiny. Accordingly, the record need not contain evidence of “bad faith, ill will or any evil motive on the part of the [recipient].” Williams v. City of Dothan, 745 F.2d 1406, 1414 (11th Cir. 1984).
This section discusses a variety of methods of proof to consider when evaluating recipient behavior to determine whether it meets the legal standard for intentional discrimination. A method of proof—or analytical framework—is an established way of organizing the evidence in an investigation or lawsuit in order to show why that evidence amounts to intentional discrimination.
Those methods are as follows:
Methods that focus on direct evidence
- Express classifications. Express classifications are the clearest form of direct evidence of discriminatory intent. If a recipient explicitly conditions the receipt of benefits or services on the race, color, or national origin of the beneficiary, or directs adverse action to be taken based on race, color, or national origin, such a policy or practice constitutes an express classification. See Section B.1.a.
- Comments or conduct by decision-makers as direct evidence of intent. The direct method of proof typically involves a statement from a decision-maker that expresses a discriminatory motive. See Section B.1.b.
- The Arlington Heights mosaic of factors. This method of proof, originally developed for Equal Protection Clause cases, uses a number of different types of circumstantial evidence that, taken collectively, can demonstrate that the recipient acted, at least in part, because of race, color, or national origin. This framework is most commonly applied in cases alleging discrimination against a group. Agencies can use this method for many different types of cases, but will find it particularly useful where the complaint is about the treatment of a group, not individuals, and the investigation reveals many different kinds of evidence. Agencies should be sure to consider this method where a complaint challenges an expressly neutral practice that has an effect on a larger class defined by race, color, or national origin. For instance, a complaint alleging that a state agency adopted a new policy with the purpose of reducing the number of minority participants could be investigated using this method. See Section B.2.
The McDonnell-Douglas framework. Plaintiffs use this framework, originally developed for Title VII employment cases, to show that a defendant treated similarly situated individuals differently because of race, color, or national origin. The framework is most commonly applied in cases alleging discrimination in individual instances. Agencies should consider using this method for investigations involving the selection of individuals, such as for program participation, benefits, or services, particularly where the recipient provides a nondiscriminatory explanation for its decision. This method is most likely to be helpful where the complaint is about one or a few individuals, and involves easily identifiable similarly situated individuals not in the protected class. For instance, a complaint alleging that a state agency denied benefits to a family because of that family’s national origin might be investigated using this method. See Section B.3.
More than one type of analysis may apply to facts disclosed in an investigation or trial to determine race-based intent. Agencies and plaintiffs can use them individually or together and may combine both direct and circumstantial evidence. Ultimately, the “totality of the relevant facts” will determine whether the recipient has engaged in intentional discrimination in violation of Title VI. See Washington v. Davis, 426 U.S. 229, 242 (1976) (discussing analysis of intentional discrimination generally).
Regardless of the method or methods of proof ultimately employed, the central question remains whether the recipient acted intentionally based on race, color, or national origin. In evaluating the totality of relevant facts, courts and federal funding agencies look to either direct or circumstantial evidence to establish whether a recipient engaged in intentional discrimination. Often, the available proof consists of a combination of these different kinds of evidence, and therefore more than one method of proof may be appropriate. The box below cross-references the major types of evidence with the related methods of proof discussed in this section.
TYPES OF EVIDENCE
Direct evidence. Direct evidence often involves a statement from a decision-maker that expresses a discriminatory motive. Direct evidence can also include express or admitted classifications, in which a recipient explicitly distributes benefits or burdens based on race, color, or national origin. Other than instances where a recipient uses race expressly to achieve diversity or implement a race-based remedy for past discrimination, finding direct evidence is rare; most recipients are circumspect enough to avoid making overtly discriminatory statements. As a result, most Title VI litigation and administrative investigations focus on circumstantial evidence. See methods of proof discussed in Section B.1.
Circumstantial evidence. Circumstantial evidence, also known as indirect evidence, requires the fact finder to make an inference or presumption. Hamilton v. Southland Christian Sch., Inc., 680 F.3d 1316, 1320 (11th Cir. 2012). “Circumstantial evidence can include suspicious timing, inappropriate remarks, and comparative evidence of systematically more favorable treatment toward similarly situated [individuals] not sharing the protected characteristic….” Loyd v. Phillips Bros., Inc., 25 F.3d 518, 522 (7th Cir. 1994); accord Troupe v. May Dep’t Stores Co., 20 F.3d 734, 736 (7th Cir. 1994). See methods of proof discussed in Sections B.2 and B.3.
Statistical evidence. Statistical evidence can often be critical in a case where the exercise of race-based motive is alleged. A plaintiff or agency investigation can use statistics in several ways to establish a claim of intentional discrimination. For example, statistics can be used show that an ostensibly race-neutral action actually causes a pattern of discrimination, a racially disproportionate impact, or foreseeably discriminatory results. While statistical evidence is not required to demonstrate intentional discrimination, plaintiffs often successfully use statistics to support, along with other types of evidence, a claim of intentional discrimination. See methods of proof discussed in Sections B.2 and C1.
Finally, it is important for agencies to remember that even if a recipient is found to have engaged in the intentional consideration of race, color, or national origin, this is not the end of the inquiry. Some uses of race are permissible. This is discussed more extensively beginning at page 30.
Title VI case law has traditionally borrowed jurisprudence from other civil rights laws with a similar structure and purpose. The remainder of this section examines methods of proving intentional discrimination in greater detail, with reference to case law not only under Title VI and the Equal Protection Clause, but also under Title VII; Title IX of the Education Amendments of 1972, 20 U.S.C. § 1681 et seq.; and Section 504 of the Rehabilitation Act of 1973, 29 U.S.C. §701, among other laws. Importantly, the analyses under these civil rights laws are not always the same, but this discussion identifies principles that are applicable to Title VI.
1. Direct Evidence of Discriminatory Intent
Direct evidence of discriminatory intent is evidence that, “if believed, proves the fact [of discriminatory intent] without inference or presumption.” Coghlan v. Am. Seafoods Co., 413 F.3d 1090, 1095 (9th Cir. 2005) (citation omitted).
Occasionally, a recipient official admits to having considered race during the decisional process as a basis for its action. In other instances, a recipient explicitly conditions the receipt of benefits or services on the race, color, or national origin of the beneficiary, or explicitly directs action be taken based on race, color, or national origin. These kinds of requirements are often referred to as “express classifications,” and are the clearest form of direct evidence.
Short of an express classification, other direct evidence of discrimination includes “any statement or document which shows on its face that an improper criterion served as the basis … for [an] adverse … action.” Fabela v. Socorro Indep. Sch. Dist., 329 F.3d 409, 415 (5th Cir. 2003). On the other hand, “remarks by non-decisionmakers or remarks unrelated to the decision making process itself are not direct evidence of discrimination.” Standard v. A.B.E.L. Servs., Inc., 161 F.3d 1318, 1330 (11th Cir. 1998).
a. Express classifications
The Equal Protection Clause requires strict scrutiny of any government policy or practice that classifies individuals based on race, color, or national origin. Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 720 (2007) (“[W]hen the government distributes burdens or benefits on the basis of individual racial classifications, that action is reviewed under strict scrutiny.”); Gratz v. Bollinger, 539 U.S. 244, 270 (2003) (applying strict scrutiny to student admissions policies that considered race as a factor). Similarly, Title VI requires recipients to demonstrate that any intentional use of race, color, or national origin classification is “narrowly tailored” to achieve a “compelling” government interest. Parents Involved, 551 U.S. at 720.
A recipient’s express or admitted use of a classification based on race, color, or national origin establishes intent without regard to the decision-makers’ animus or ultimate objective. Such classifications demonstrate a discriminatory purpose as a matter of law. See Miller v. Johnson, 515 U.S. 900, 904–05 (1995); see also Wittmer v. Peters, 904 F. Supp. 845, 849–50 (C.D. Ill. 1995), aff’d, 87 F.3d 916 (7th Cir. 1996). “Put another way, direct evidence of intent is ‘supplied by the policy itself.’” Hassan v. City of New York, 804 F.3d. 277, 295 (3d Cir. 2015) (quoting Massarsky v. Gen. Motors Corp., 706 F.2d 111, 128 (3d Cir.1983) (Sloviter, J., dissenting)).
Where a plaintiff demonstrates, or an agency determines, that a challenged policy overtly and expressly singles out a protected group for disparate treatment, “a plaintiff need not prove the malice or discriminatory animus of a defendant ….” Bangerter v. Orem City Corp., 46 F.3d 1491, 1501 (10th Cir. 1995); see also Ferrill v. Parker Grp., Inc., 168 F.3d 468, 473 n.7 (11th Cir. 1999) (“[I]ll will, enmity, or hostility are not prerequisites of intentional discrimination.”). Rather, the focus is on the “explicit terms of the discrimination,” Int’l Union, United Auto. Aerospace & Agric. Implement Workers of Am. v. Johnson Controls, Inc., 499 U.S. 187, 199 (1991); that is, how the recipient’s actions specifically deprived or otherwise adversely affected the individual or individuals of access to a federally funded program or benefit. Even benign motivations for racial classifications are presumptively invalid and trigger strict scrutiny in Equal Protection Clause and Title VI cases. Adarand, 515 U.S. at 223–24 (1995); Grutter, 539 U.S. at 326.
Even without a direct admission or express policy, a plaintiff may prove intentional discrimination with other forms of direct evidence demonstrating that the “decisionmakers placed substantial negative reliance on an illegitimate criterion in reaching their decision.” Price Waterhouse v. Hopkins, 490 U.S. 228, 277 (1989) (O’Connor, J., concurring);  Venters v. City of Delphi, 123 F.3d 956, 972 (7th Cir. 1997) (direct evidence includes “evidence which in and of itself suggests” that someone with managerial authority was “animated by an illegal ... criterion.”). For example, a statement of an official involved in the decision stating that an ostensibly race-neutral action was taken in order to limit minority individuals’ eligibility for a federally funded benefit or program is direct evidence of race-based intent. Even isolated comments may constitute direct evidence of discrimination if they are “contemporaneous with the [adverse action] or causally related to the [adverse action] decision making process.” Kennedy v. Schoenberg, Fisher & Newman, Ltd., 140 F.3d 716, 723 (7th Cir. 1998) (citations omitted).
This type of direct evidence of discriminatory intent does not require “a virtual admission of illegality.” Venters, 123 F.3d at 973. For example, direct evidence need not take the form of an admission where the defendant states “I’m [taking this adverse action] because you’re in a protected group.” Sheehan v. Donlen Corp., 173 F.3d 1039, 1044 (7th Cir. 1999); see Venters, 123 F.3d at 973. The court in Venters explained that “the evidence need not be this obvious to qualify as direct evidence.” Id. And the Sheehan court explained why: because such a requirement “would cripple enforcement of the ... discrimination laws.” Sheehan, 173 F.3d at 1044. The direct evidence of such remarks must, however, establish that race was an important factor motivating the challenged action. “Stray remarks,” “derogatory comments,” even those uttered by decision-makers, may not constitute direct evidence of discrimination if unrelated to the adverse decision. Price Waterhouse, 490 U.S. at 277 (O’Connor, J., concurring); Fuentes v. Perskie, 32 F.3d 759, 767 (3d Cir. 1994). Evidence of such remarks or comments is nevertheless important in an intent case, and can help to establish circumstantial or indirect evidence of intent. Doe v. C.A.R.S. Prot. Plus, Inc., 527 F.3d 358, 368 (3d Cir. 2008); Fitzgerald v. Action, Inc., 521 F.3d 867, 877 (8th Cir. 2008) (same); see also Lounds v. Lincare, Inc., 812 F.3d 1208, 1224 (10th Cir. 2015) (citing Kerri Lynn Stone, Taking in Strays: A Critique of the Stray Comment Doctrine in Employment Discrimination Law, 77 Mo. L. Rev. 149, 177 (2012) (“[S]tray remarks can prove to be invaluable insights into biases at every level of consciousness that may be rife but invisible within the workplace.... [They] may bespeak a workplace culture in which certain language or sentiments are tolerated and perhaps encouraged or rewarded.”)).
By way of illustration, in Wilson v. Susquehanna Township Police Dep’t, 55 F.3d 126 (3d Cir. 1995), a Title VII case, a female plaintiff alleged that she was not promoted because of her sex. The plaintiff’s evidence revealed a number of discriminatory occurrences, including the daily circulation of sexually explicit drawings, the posting of obscene notices (some referring to female employees by name), sexual conversations between officers and female employees, the showing of an x-rated movie and graphic home videos in the station house, the Chief’s regular discussion of sex lives and employees’ anatomy, the Chief’s bemused dismissal of the plaintiff’s complaint about an indecent assault committed by an officer, and the Chief’s comment that he did not promote the plaintiff because the town manager “wanted a man.” Id. at 127–29. The court of appeals described that evidence as direct evidence of intentional sex discrimination, explaining that “[t]he record clearly goes beyond ‘stray remarks’ and evinces strong gender bias in the police department.... This evidence, which included ‘conduct or statements by persons involved directly reflecting the discriminatory attitude,’ ... constitutes ‘direct evidence’ of discriminatory animus.” Id. at 130 (citations and quotations omitted).
In In re Rodriguez, 487 F.3d 1001, 1006–08 (6th Cir. 2007), a case originally brought under Michigan’s Civil Rights Act, which borrows legal standards from federal civil rights laws including Title VII,  the court found that a Hispanic employee was not selected for promotion based on a manager’s impression about the applicant’s “language” and “how he speaks.” This evidence, the court held, was direct evidence of discrimination. Stating that “the [EEOC] recognizes linguistic discrimination as national origin discrimination” and that “discriminationbased on manner of speaking can be national origin discrimination,” the court found that the plaintiff’s “Hispanic speech pattern and accent” played a motivating part in the manager’s decision to deny the plaintiff a promotion. Id. at 1008–09; accord, Diaz v. Jiten Hotel Mgmt., Inc., 762 F. Supp. 2d 319, 337 (D. Mass. 2011) (“racially, sexually, or ageist offensive language is necessarily prejudicial, precisely because it is highly probative”).
A clean “direct evidence” case—where direct evidence alone establishes that discrimination was the sole reason for an adverse decision—is rare. Price Waterhouse, 490 U.S. at 271 (“[D]irect evidence of intentional discrimination is hard to come by.”) (O’Connor, J., concurring). After all, decision-makers seldom will admit that they based decisions on race or ethnic origin, or used either as a criterion. See, e.g., SECSYS, LLC v. Vigil, 666 F.3d 678, 686 (10th Cir. 2012).
2. The Arlington Heights Framework
Many cases of intentional discrimination are not proven by a single type of evidence. Rather, many different kinds of evidence-direct and circumstantial, statistical and anecdotal-are relevant to the showing of intent and should be assessed on a cumulative basis. Arlington Heights, 429 U.S. at 266–68, and its progeny set forth a variety of factors probative of intent to discriminate. Under this method of proving intent, the court or investigating agency analyzes whether discriminatory purpose motivated a recipient’s actions by examining factors such as statistics demonstrating a “clear pattern unexplainable on grounds other than” discriminatory ones; “[T]he historical background of the decision”; “[T]he specific sequence of events leading up to the challenged decision”; the defendant’s departures from its normal procedures or substantive conclusions, and the relevant “legislative or administrative history.” Faith Action for Cmty. Equity v. Hawai’i, No. CIV. 13-00450 SOM, 2015 WL 751134, at *7 (D. Haw. Feb. 23, 2015) (Title VI case citing Pac. Shores Props., LLC v. City of Newport Beach, 730 F.3d 1142, 1158–59 (9th Cir. 2013)); see also Sylvia Dev. Corp. v. Calvert Cty., 48 F.3d 810, 819 (4th Cir. 1995) (adding to the Arlington Heights factors evidence of a “consistent pattern” of actions of decision-makers that have a much greater harm on minorities than on non- minorities). When a recipient applies different procedural processes or substantive standards to requests of minorities and non-minorities, the use of such different processes or standards, when a non-minority receives more favorable treatment, may raise an inference of discriminatory intent. “These factors are non-exhaustive.” Pac. Shores Props., 730 F.3d at 1159.
AGENCY PRACTICE TIP
Agencies can use the Arlington Heights framework for many different types of cases, but will find it particularly useful where the complaint is about the treatment of a group, not individuals, and the investigation reveals many different kinds of evidence. Agencies should be sure to consider this method where a complaint challenges an expressly neutral policy or practice that has an effect on a larger class defined by race, color, or national origin. For instance, an agency could use this method when investigating a complaint alleging that a state agency adopted a new policy with the purpose of reducing the number of minority participants.
In court and agency investigations, evaluation of these factors “demands a sensitive inquiry into such circumstantial and direct evidence of intent as may be available.” Arlington Heights, 429U.S. at 266. Moreover, when a plaintiff relies on the Arlington Heights method to establish intent, “the plaintiff need provide very little such evidence ... to raise a genuine issue of fact ...; any indication of discriminatory motive ... may suffice to raise a question that can only be resolved by a fact-finder.” Pac. Shores Props., 730 F.3d at 1159 (citations omitted).
FACTORS/CIRCUMSTANTIAL EVIDENCE PROBATIVE OF INTENT
- Statistics demonstrating a clear pattern of discriminatory effect;
- The historical background of the decision and other decisions on comparable matters;
- The sequence of events leading up to the decision, as compared to other decisions on comparable matters;
- Departures from normal procedures or substantive conclusions;
- Relevant legislative or administrative history; and
- Consistent pattern of actions of decision-makers that impose much greater harm on minorities than on non-minorities.
Critically, Arlington Heights directs courts and agencies to engage in a cumulative assessment of the evidence. By way of illustration, in North Carolina State Conference of NAACP v. McCrory, No. 1:13CV658, 2016 WL 1650774, at *5 (M.D.N.C. Apr. 25, 2016), plaintiffs challenged provisions of a North Carolina election law, alleging that discriminatory intent to disenfranchise African-American voters motivated the legislature in violation of the Fourteenth and Fifteenth Amendments and the Voting Rights Act. The Fourth Circuit agreed. N.C. State Conf. of NAACP v. McCrory, 831 F.3d 204 (4th Cir. 2016). The district court’s error in holding otherwise, the Fourth Circuit explained, “resulted from the court’s consideration of each piece of evidence in a vacuum, rather than engaging in the totality of the circumstances analysis required by Arlington Heights.” Id. at 233. The district court “missed the forest in carefully surveying the many trees.” Id. at 214. Instead, agencies evaluating possible intentional discrimination by recipients must conduct a cumulative assessment of all the available evidence.
This case also illustrates the kinds of evidence relevant to each of the Arlington Heights factors described above:
- Historical background of the decision. First, the court considered the historical background in the state generally and related to voting in particular, identifying “North Carolina’s history of race discrimination and recent patterns of official discrimination, combined with the racial polarization of politics in the state” as particularly relevant. Id. at 223. Against this background of historical discrimination in the state, the court found “the record is replete with evidence of instances since the 1980s in which the North Carolina legislature has attempted to dilute the voting rights of African Americans” and pointed to the numerous instances of “Department of Justice and federal court determinations have determined that the North Carolina General Assembly acted with discriminatory intent .…” Id. The court found these examples revealed “a series of official actions taken for invidious purposes,” and held that the district court “erred in minimizing these facts.” Id. (citing Arlington Heights, 429 U.S. at 267).
- Sequence of events leading to the decision. Next, the court turned to an examination of the sequence of events leading to the legislature’s passage of the challenged provisions, finding these events “devastating” to the defense. N.C. State Conf. of NAACP, 831 F.3d at 227. The court found that the undisputed sequence of events—“the General Assembly’s eagerness to … rush through the legislative process the most restrictive voting law North Carolina has seen since the era of Jim Crow—bespeaks a certain purpose …. Although this factor, as with the other Arlington Heights factors, is not dispositive on its own, it provides another compelling piece of the puzzle of the General Assembly’s motivation.” Id. at 229.
- Legislative history leading to the decision. As instructed by Arlington Heights, the court also considered the sequence of events described above from the perspective of “legislative history” because such evidence “may be highly relevant, especially where there are contemporaneous statements by members of the decisionmaking body, minutes of its meetings, or reports.” Id. (citing Arlington Heights, 429 U.S. at 268). The record revealed that the General Assembly requested a report on voting patterns, and that data established that African Americans in North Carolina disproportionately used early voting, same-day registration, and out-of-precinct voting. N.C. State Conf. of NAACP, 831 F.3d at 230. The court held that “relying on this data, the General Assembly enacted legislation restricting all—and only—practices disproportionately used by African Americans …. [W]e cannot ignore the choices the General Assembly made with this data in hand.” Id.
Impact. The first Arlington Heights factor, statistics demonstrating a clear pattern of discriminatory effect, acknowledges that disparate impact evidence can be probative of discriminatory intent. Arlington Heights, 429 U.S. at 266 (discussing the importance of the impact of the official action, including “whether it bears more heavily on one race than another”). Here, the court analyzed the available impact data and held that the same data showing that African Americans disproportionately used each of the voting mechanisms removed by the new provisions also established “sufficient disproportionate impact” for an Arlington Heights analysis. N.C. State Conf. of NAACP, 831 F.3d at 231.
The court conducted a cumulative assessment of this evidence:
[T]he totality of the circumstances—North Carolina’s history of voting discrimination; the surge in African American voting; the legislature’s knowledge that African Americans voting translated into support for one party; and the swift elimination of the tools African Americans had used to vote and imposition of a new barrier at the first opportunity to do so—cumulatively and unmistakably reveal that the General Assembly used [the new law] to entrench itself.
Id at 233. Accordingly, when viewed collectively, the evidence in the record established intentional discrimination based on race. Id.
Finally, it is important to understand that under the Arlington Heights framework, evidence identifying similarly situated comparators is helpful but not required. In this regard, the relationship between the Arlington Heights framework and the McDonnell-Douglas framework is sometimes misunderstood. As discussed more extensively below in Section B.3., the McDonnell-Douglas method of proof requires a showing that the recipient treated one or a few similarly situated individuals differently because of race, color, or national origin. However, plaintiffs alleging intentional discrimination under civil rights statutes “need not demonstrate the existence of a similarly situated entity who or which was treated better than the plaintiff in order to prevail.” Pac. Shores Props., 730 F.3d at 1158-59 (explaining that a plaintiff need not rely on the McDonnell-Douglas approach to intentional discrimination but may instead produce circumstantial evidence of intentional discrimination using the Arlington Heights method). McDonnell Douglas “is not a straightjacket requiring the plaintiff to demonstrate that such similarly situated entities exist” but is just one way to prove intentional discrimination. Id. at 1159.
Impact evidence. In many cases, including many litigated under Arlington Heights, evidence will show that an ostensibly race-neutral practice has had a much more harmful effect on minorities than on non-minorities. Arlington Heights instructs courts and agencies to consider “the impact of the official action” including whether “it bears more heavily on one race than another.” 429 U.S. at 266 (citations and quotations omitted). Accordingly, the discriminatory impact of a facially neutral policy or practice (frequently, but not always, demonstrated through the use of statistics) can be used as part of the evidentiary showing in an intentional discrimination case. See Melendres v. Arpaio, 989 F. Supp. 2d 822, 902 (D. Ariz. 2013) (awarding injunctive relief to Title VI plaintiffs and finding that plaintiffs demonstrated “racially disparate results” and “additional indicia of discriminatory intent”) (citing Feeney, 442 U.S. at 272); see also Arlington Heights, 429 U.S. at 264–66; Comm. Concerning Cmty. Improvement v. City of Modesto, 583 F.3d 690 (9th Cir. 2009) (Title VI and equal protection case finding that statistical evidence was sufficient to create inference of intent where race-neutral precondition to receiving municipal services served to exclude Latino-majority neighborhoods)).
In only rare instances will a showing of disparate impact by itself support a showing of discriminatory intent¾for example, where racially variant results cannot be explained on other grounds, such as in cases of a dramatic mismatch between jury representation and the composition of a surrounding community. Castaneda v. Partida, 430 U.S. 482, 495–96 (1977). In most instances, however, “impact alone is not determinative, and the Court must look to other evidence.” Arlington Heights, 429 U.S. at 266, 267–68 (enumerating factors that indicate evidence of intent) (footnotes omitted).
When attempting to rely on impact evidence in an intent case, the plaintiff must, as an initial matter, precisely identify the “facially neutral policy or practice” at the heart of the discrimination claim. (The Title VI Legal Manual’s disparate impact section discusses this requirement in detail.) In addition, in Arlington Heights, the selection of a similarly situated comparator group is a key feature of cases where plaintiffs proffer impact evidence. By its nature, “disparate impact” evidence involves showing a disparity. Plaintiff must show that the extent of harm the policy or practice causes minorities and non-minorities is different. The level or degree of impact that a plaintiff alleging discriminatory intent must show depends on a variety of factors, including the strength of the impact evidence and the strength of other indicators of intent under Arlington Heights. But, as one court noted, “[i]t would be improper to posit a quantitative threshold above which statistical evidence of disparate racial impact is sufficient as a matter of law to infer discriminatory intent, and below which it is insufficient as a matter of law.” Gay v. Waiters’ & Dairy Lunchmen’s Union, Local No. 30, 694 F.2d 531, 551 (9th Cir. 1982). Because disparate impact is not the only factor in an Arlington Heights case, “showing disproportionate impact, even if not overwhelming impact, suffices to establish one of the circumstances evidencing discriminatory intent.” N. Carolina State Conference of NAACP, 831 F.3d at 231.
In addition, impact evidence most often involves the presentation of statistical evidence. Thomas v. Washington Cty. Sch. Bd., 915 F.2d 922, 926 (4th Cir. 1990). However, statistical evidence, while extremely beneficial, is not a necessity in impact cases. Id. Indeed, a series of “discrete episodes” negatively affecting minorities can raise a plausible inference of discriminatory impact. McCoy v. Canterbury, No. 3:10-0368, 2010 WL 5343298, at *5 (S.D.W. Va. Dec. 20, 2010), aff’d, 428 Fed. App’x 247 (4th Cir. 2011). Accordingly, non-statistical evidence of harm to minorities and non-minorities that is significantly different will be relevant evidence in an Arlington Heights case.
Moreover, statistics alone will seldom prove discriminatory intent. There may be cases where statistics establish “a clear pattern, unexplainable on grounds other than race,” “but such cases are rare.” Arlington Heights, 429 U.S. at 266, No matter how “devastating or reliable” the statistics appear to be, Ward v. Westland Plastics, Inc., 651 F.2d 1266, 1270 (9th Cir. 1980) (per curiam), they must reveal that some “invidious discriminatory purpose” is causing the disparate outcomes. Arlington Heights, 429 U.S. at 266; see also Feeney, 442 U.S. at 279 (plaintiff must show that the rule was promulgated or reaffirmed “‘because of,’ not merely ‘in spite of,’ its adverse impact on” persons in the plaintiff’s class); Horner v. Ky. High Sch. Athletic Ass’n, 43 F.3d 265, 276 (6th Cir. 1994) (citing Feeney). As such, and in most instances, “the question whether the facts proved are sufficient to permit a legal inference of discriminatory intent cannot properly be reduced into a mere battle of statistics.” Gay, 694 F.2d at 552. Absent a “stark” pattern, then, discriminatory intent requires more than discriminatory impact. Arlington Heights, 429 U.S. at 266.
Recipient’s awareness of the impact. Also consistent with the Arlington Heights factors is an inquiry into whether the discriminatory impact of the challenged action was foreseeable:
[A]ctions having foreseeable and anticipated disparate impact are relevant evidence to prove the ultimate fact, forbidden purpose.... [T]he foreseeable effects standard [may be] utilized as one of the several kinds of proofs from which an inference of segregative intent may be properly drawn.... Adherence to a particular policy or practice, with full knowledge of the predictable effects of such adherence ... is one factor among many others which may be considered by a court in determining whether an inference of segregative intent should be drawn.
Columbus Bd. of Educ. v. Penick, 443 U.S. 449, 464–65 (1979); see United States v. Brown, 561 F.3d 420, 433 (5th Cir. 2009). Foreseeability is a common feature of Title VI and equal protection claims, and allegations that properly package foreseeability together with factors such as impact and history of defendant’s actions, have succeeded. See, e.g., N.C. State Conf. of NAACP, 831 F.3d at 223; Dowdell v. City of Apopka, 698 F.2d 1181, 1186 (11th Cir. 1983) (discussing “obviously foreseeable” outcome of the town’s decision to spend nearly all of its revenue-sharing monies on the white community, at the expense of communities of color); United States v. Bannister, 786 F. Supp. 2d 617, 665–66 (E.D.N.Y. 2010) (expressing support for using discriminatory impact, foreseeable consequences, and historical background to demonstrate intent in enacting mandatory minimums for crack cocaine, but determining that court could not find intentional discrimination where Second Circuit already made finding on the specific issue under consideration).
Additional examples of successful outcomes where impact and foreseeable consequences combine with other Arlington Heights factors, such as history of state action, include the following:
- Spanish-speaking food stamp beneficiaries alleged that state agencies administering the state food stamp program continued a policy of failing to ensure bilingual services for food stamp applicants who were limited English proficient. The plaintiffs alleged that the defendants continued this policy while knowing that Spanish-speaking applicants and beneficiaries were being harmed as a consequence. The court found that such knowledge was sufficient to state a Title VI claim that the defendants purposefully acted based on national origin, finding that “disparate impact, history of the state action, and foreseeability and knowledge of the discriminatory onus placed upon the complainants” is the type of circumstantial evidence upon which a case of intentional discrimination is often based. Almendares v. Palmer, 284 F. Supp. 2d 799, 806 (N.D. Ohio 2003) (citations omitted)
- A facially neutral NCAA rule (Proposition 16) raising the minimum academic requirements for incoming college athletes to qualify for athletic scholarships and compete in college sports applied to all incoming college athletes but had a statistically greater adverse impact on black athletes. The NCAA was aware that the impact of the proposed rule would reduce the number of black athletes qualifying for athletic scholarships, and adopted the rule specifically to promote higher academic standards among black athletes. The court held that plaintiffs had stated a claim of purposeful discrimination under Title VI. Pryor v. NCAA, 288 F.3d 548, 562 (3d Cir. 2002). Pryor directly addressed the Arlington Heights standards for intentional discrimination, concluding that the plaintiffs met the intent test where the NCAA had actual notice and knowledge of the impact on black athletes, and affirmatively considered that impact in reaching its decision to adopt Proposition 16.
- Plaintiffs claimed intentional discrimination based partly on the defendant’s knowledge of the impact that placement of a cement grinding facility would have on the minority community, together with allegations regarding historical practices and a specific sequence of events leading to the placement decision. The court found that the plaintiffs “not only showed that the operation of the cement grinding facility would have a disparate impact upon the predominantly minority community … but also that the [defendant] was well-aware of the potential disproportionate and discriminatory burden placed upon that community and failed to take measures to assuage that burden.” The court further determined that the plaintiffs had stated a claim of intentional discrimination under Title VI, sufficient to survive the defendant’s motion to dismiss. The court set forth that “the controlling decisions of the Supreme Court and the Third Circuit make it clear that a case of intentional discrimination is often based upon the type of circumstantial evidence which the … Plaintiffs allege …, namely, disparate impact, history of the state action, and foreseeability and knowledge of the discriminatory onus placed upon the complainants.” S. Camden, 254 F. Supp. at 496–97 (citing Arlington Heights, 429 U.S. at 267; Penick, 443 U.S. at 465 (1979); Pryor, 288 F.3d at 563).
3. The McDonnell-Douglas Framework
Another common way to prove intentional discrimination is to establish that a recipient treated similarly situated individuals differently because of race, color, or national origin.
1) Step 1—The prima facie case
Plaintiff must first prove a prima facie case of discrimination by a preponderance of the evidence. To establish a prima facie case of intentional discrimination under Title VI using the McDonnell-Douglas framework from Title VII, a plaintiff typically shows that he or she is a member of a particular protected group, was eligible for the recipient’s program, activity or service, and was not accepted into that program or otherwise treated in an adverse manner, and that an individual who was similarly situated with respect to qualifications, but was not in the plaintiff’s protected group was given better treatment. See, e.g., Brewer v. Bd. of Trs. of Univ. of Ill., 479 F.3d 908, 921 (7th Cir. 2007) (Title VI case where court found that plaintiff’s case “falls apart because of a failure to locate a similarly situated individual”).
AGENCY PRACTICE TIP
Agencies can use the McDonnell-Douglas framework for investigations involving the selection of individuals, such as for program participation, benefits, or services, particularly where the recipient provides a nondiscriminatory explanation for its decision. This method is most likely to be helpful where the complaint is about one or a few individuals, and involves easily identifiable similarly situated individuals not in the protected class. For instance, a complaint alleging that a state agency denied benefits to a family because of that family’s national origin might be investigated using this method.
With respect to what constitutes adverse action or “harm,” there are “no bright-line rules,” Wanamaker v. Columbian Rope Co., 108 F.3d 462, 466 (2d Cir. 1997), so courts and agencies must make that determination in each case. As such, whether conduct rises to the level of “adverse action” is a fact-specific inquiry. The harm need not be physical in nature, or even the type of harm that would permit an award of compensatory damages. For example, the Supreme Court has held that intentional racial segregation is a harm in and of itself. See Brown v. Bd. of Educ., 347 U.S. 483 (1954). Similarly, the stigma that intentional discrimination may cause is a cognizable harm. See generally Johnson v. California, 543 U.S. 499, 507 (2005) (“racial classifications ‘threaten to stigmatize individuals by reason of their membership in a racial group’”) (quoting Shaw v. Reno, 509 U.S. 630, 643 (1993)). The provision of fewer or inferior services or benefits to a person or class of persons will satisfy the adversity requirement, but adversity can be established even without the loss of specific services or benefits; threatened or imminent harm can satisfy the adverse action requirement.
Moreover, Title VI’s broad nondiscrimination mandate means that investigating agencies generally should take an inclusive approach to determining legally sufficient harms. Title VI’s plain language supports this approach. The statute states that no person shall on the ground of race, color, or national origin “be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.” 42 U.S.C. § 2000d. Agency regulations further state that recipients may not administer their programs or activities in a manner that “den[ies] any individual any disposition, service, financial aid, or benefit provided under the program,” 28 C.F.R. § 42.104(b)(1)(i) (DOJ) (emphasis added), or “restrict[s] an individual in any way in the enjoyment of any advantage or privilege enjoyed by others receiving any disposition, service, financial aid, or benefit under the program,” Id. § 42.104(b)(1)(iv) (emphasis added). This language is best read to encompass a broad range of “adverse actions” that may be caused by a recipient’s administration of its program.
2) Step 2 – The defendant must articulate a legitimate non-discriminatory reason
If the plaintiff establishes a prima facie case, the burden in court shifts to the defendant to articulate some legitimate, nondiscriminatory reason for the challenged action. EEOC v. Boeing Co., 577 F.3d 1044, 1049 (9th Cir. 2009). The defendant’s explanation of its legitimate reasons must be clear and reasonably specific; not all proffered reasons would be legally sufficient to rebut a prima facie case. See Texas Dep’t of Cmty. Affairs v. Burdine, 450 U.S. 248, 254–55, 258 (1981). For example, in the employment context, a defendant may not merely state that the employment decision was based on the hiring of the “best qualified” applicant, but must provide specifics regarding that applicant’s qualifications, such as seniority, length of service in the same position, personal characteristics, general education, or experience in comparable work, and must demonstrate why that person’s qualifications were considered superior to those of the plaintiff. See Steger v. Gen. Elec. Co., 318 F.3d 1066, 1075–76 (11th Cir. 2003).
3) Step 3 – The plaintiff must demonstrate pretext
If the defendant meets the Step 2 burden, the burden shifts back to the plaintiff to demonstrate that the proffered reason is false—that is, that the nondiscriminatory reason(s) the defendant gives for its actions are not the true reasons and are actually a pretext for the exercise of prohibited discriminatory intent. Brooks v. Cty. Comm’n of Jefferson Cty., 446 F.3d 1160, 1162– 63 (11th Cir. 2006) (addressing a Title VII race discrimination claim). A plaintiff can show pretext by pointing to “weaknesses, implausibilities, inconsistencies, incoherencies, or contradictions” in the defendant’s proffered legitimate reasons for its action, such that a reasonable fact finder could rationally find them unworthy of credence. Id. at 1163 (quoting Jackson v. Ala. State Tenure Comm’n, 405 F.3d 1276, 1289 (11th Cir. 2005)); Mickelson v. N.Y. Life Ins. Co., 460 F.3d 1304, 1315 (10th Cir. 2006). Plaintiffs can, for example, present evidence that the defendant’s stated reasons for taking the adverse action were false; the defendant acted contrary to a written policy setting forth the action the defendant should have taken under the circumstances; or the defendant acted contrary to an unwritten policy or practice when making the decision. See Plotke v. White, 405 F.3d 1092, 1102 (10th Cir. 2005). A plaintiff may also show pretext through evidence that the “employer’s proffered non-discriminatory reasons [were] either a post hoc fabrication or otherwise did not actually motivate the employment action ….” Fuentes, 32 F.3d at 764.
AGENCY PRACTICE TIP
As mentioned previously, certain procedural aspects of the methods of proof developed in the litigation context do not transfer to the administrative context. Here, the McDonnell-Douglas burden- shifting test that applies in litigation to determine whether an institution has engaged in intentional discrimination does not necessarily apply in the context of agency enforcement activities prior to administrative litigation. An agency is free to collect and analyze the evidence described in the steps below as part of its initial investigation, or may choose to make a preliminary prima facie finding and require the recipient to articulate its defense as a next step.
The Supreme Court has cautioned that the four McDonnell-Douglas elements are not “an inflexible formulation.” Teamsters, 431 U.S. at 358. Further, as previously noted, agency Title VI investigations generally follow a non-adversarial model that does not involved burden- shifting. Nevertheless the McDonnell-Douglas framework may be useful for complaint investigations, particularly where the investigation uncovers evidence of similarly situated comparators who were treated differently or better. The example below, from joint DOJ and Department of Education guidance, illustrates how the McDonnell-Douglas framework would inform an administrative investigation.
ILLUSTRATION: MCDONNELL DOUGLAS FRAMEWORK APPLIED TO INVESTIGATION OF ALLEGED DISCRIMINATORY SCHOOL DISCIPLINE
Complaint. Plaintiff alleged discrimination after a school imposed different disciplinary sanctions on two students in the sixth grade—a non-Hispanic student and a Hispanic student—who engaged in a fight. Both students had similar disciplinary histories, having each previously received after-school detention for minor infractions. The Hispanic student received a three-day out-of-school suspension for the student’s involvement in the fight, while the non-Hispanic student received a two-day out-of-school suspension for the same misconduct, raising a concern that the students were treated differently based on race.
Based on these facts and circumstances, the Departments of Education and Justice would make an initial determination that the students were similarly situated, as they were involved in the same incident and have similar discipline records. If the school provided evidence of facts and circumstances surrounding the incident that would constitute a legitimate, nondiscriminatory reason for the different treatment, such as evidence that it disciplined the Hispanic student more severely because the student instigated the fight and directly threatened school officials who tried to break up the fight, then these facts and circumstances might constitute a nondiscriminatory reason for the different treatment. If the school failed to provide a legitimate nondiscriminatory reason for imposing a different sanction on either student, the Departments could find that the school had violated Title VI.
If, however, the school did provide a legitimate, nondiscriminatory reason for the different sanction, the Departments would probe further to determine whether the reason given for the enhanced sanction was an accurate statement of the reasons for different treatment of the two students, or constituted a pretext for racial discrimination. In making this determination, the Departments would request and consider information such as witness statements, codes of conduct, and student disciplinary records. The Departments would then evaluate, among other things, whether the school conformed to its written policies; whether the Hispanic student did, in fact, instigate the fight; and whether the school had previously imposed a higher sanction on non-Hispanic students who had instigated fights.
1. Proof of Systemic or Wide-Spread Discrimination (Pattern or Practice Discrimination)
Principles similar to those discussed above may be used to establish that a recipient engaged in widespread discrimination in violation of Title VI. In these cases, one means of proving intentional discrimination is through circumstantial evidence showing a statistical disparity that affects a large number of individuals. Agencies investigating complaints alleging widespread discrimination may find useful guidance in Title VII case law that discusses “pattern or practice” discrimination. The phrase “pattern or practice” can be used to describe a systemic violation of Title VI, regardless of the method of proof employed. Although statistical evidence is usually used to establish a pattern or practice of intentional discrimination, it is not required to establish wide-spread or systemic discrimination. This section focuses on the use of statistical evidence of disparity to establish a pattern showing different treatment based on race, color, or national origin.
In International Brotherhood of Teamsters v. United States, 431 U.S. 324 (1977), a case brought under the “pattern or practice” provision of Title VII, the Court stated that “statistics showing racial or ethnic imbalance are probative … because such imbalance is often a telltale sign of purposeful discrimination.” Id. at 339 n.20. Accordingly, statistical evidence of a sufficiently “gross disparity” between the affected population and the general population may establish an inference of intentional discrimination. Hazelwood Sch. Dist. v. United States, 433 U.S. 299, 307–08 (1977) (“Where gross statistical disparities can be shown, they alone may in a proper case constitute prima facie proof of a pattern or practice of discrimination.”). As previously noted, the term “pattern or practice” can be used broadly to refer to systemic discrimination. The term “pattern or practice” also refers to a technical claim type authorized by various civil rights statutes. These statutes use the term to define the authority of the Attorney General or private parties to bring certain claims in court. See, e.g., Title VII, 42 U.S.C. § 2000e-6(a); The Violent Crime Control and Law Enforcement Act of 1994, 42 U.S.C. § 14141(b); The Omnibus Crime Control and Safe Streets Act of 1968, 42 U.S.C. § 3789d(c)(3). A Title VII pattern or practice case, for example, will demonstrate that an employer is taking action that causes the same kind of harm to a great number of individuals. In Teamsters, the employer used job transfer policies that punished individuals, primarily minorities, who tried to transfer from less desirable jobs to more desirable ones. The “pattern or practice” that was challenged harmed many minorities in precisely the same manner. While Title VI does not expressly include a “pattern or practice” claim, principles developed in these contexts and discussed below can nevertheless inform the investigation and analysis of Title VI claims. See, e.g., Melendres v. Arpaio, 695 F.3d 990 (9th Cir. 2012) (class action alleging pattern or practice of racial profiling by law enforcement agency in violation of Title VI and the Fourth and Fourteenth Amendments); Dep’t of Justice, Investigation of Los Angeles County Sheriff’s Department Stations in Antelope Valley (June 28, 2013) (Title VI pattern or practice violation). 
For Title VI, that kind of widespread or broad discriminatory practice is often viewed or described as a claim of “systemic discrimination”—a practice that harms a large number of minority individuals in the same manner. For example, were a written test used to determine eligibility for a federally funded benefit or program, and the test resulted in a much higher percentage of minorities than non-minorities being determined ineligible for the benefit or access to the program, that might present a case of systemic discrimination. The method of proof used in pattern or practice cases under other statutes can be applied to these kinds of Title VI cases.
To prove such systemic discrimination using this method in a Title VI case, the plaintiff must show that discrimination was the recipient’s standard operating procedure; that is, the plaintiff must “prove more than the mere occurrence of isolated or accidental or sporadic discriminatory acts.” EEOC v. Joe’s Stone Crab, Inc., 220 F.3d 1263, 1286–87 (11th Cir. 2000) (quoting Teamsters, 431 U.S. at 336 (internal quotation marks omitted)). Rather, the plaintiff must establish by a preponderance of the evidence that discrimination is the company’s “regular rather than unusual practice.” Joe’s Stone Crab, 220 F.3d at 1287 (quoting Teamsters, 431 U.S. at 336). A plaintiff in a pattern or practice case can prove that discrimination was the defendant’s “standard operating procedure” by, among other things, presenting statistical evidence of similarly situated individuals not in the protected class who were treated better than those in the protected class. Craik v. Minn. State Univ. Bd., 731 F.2d 465, 470 (8th Cir. 1984).
In a case alleging such pervasive or systemic discrimination, the plaintiff need not initially show discrimination against any particular person; rather the critical showing at the prima facie stage is one of a pervasive policy of intentional discrimination affecting many individuals. See Teamsters, 431 U.S. at 360; Chin v. Port Auth. of N.Y. & N.J., 685 F.3d 135, 147 (2d Cir. 2012) (noting that in such cases "the government need not demonstrate specific losses to specific individuals to establis that injunctive relief is appropriate"). Once the plaintiff has established a prima facie case, the defendant can rebut it by either demonstrating that the plaintiff based his or her statistical calculations on faulty data, flawed computations, or improper methodologies, or by introducing alternative statistical evidence. Teamsters, 431 U.S. at 360 & n.46. As in other disparate treatment cases, the ultimate burden of persuasion rests with the plaintiff. Id. at 362 n.50 (citing McDonnell-Douglas, 411 U.S. at 804–06). If the defendant fails to rebut the inference that arises from the plaintiff’s prima facie case, the court can conclude “that a violation has occurred.” Id. at 361.
AGENCY PRACTICE TIP
As emphasized above in the McDonnell-Douglas discussion, certain procedural aspects of methods of proof developed in the litigation context do not transfer to the administrative context. Here, the Title VII burden-shifting test for formal “pattern or practice” claims that applies in litigation to determine whether an institution has engaged in intentional discrimination does not necessarily apply in the context of agency enforcement activities prior to litigation. An agency is free to collect and analyze all the evidence described in this section as part of its initial investigation, or may choose to make a preliminary prima facie finding and require the recipient to articulate its defense as a next step.
As previously stated, statistics typically are used to help establish that a pattern of discrimination based on race, color, or national origin was the recipient’s “standard operating procedure.” Teamsters, 431 U.S. at 336; Hazelwood, 433 U.S. at 307. Statistics showing racial or ethnic imbalance are probative in pattern or practice cases because a clear and significant imbalance based on race or ethnicity is often an indication of purposeful discrimination. Teamsters, 431 U.S. at 336; Hazelwood, 433 U.S. at 307. Statistics showing racial or ethnic imbalance are probative in pattern or practice cases because a clear and significant imbalance based on race or ethnicity is often an indication of purposeful discrimination. Teamsters, 431U.S. at 339 n.20; Lujan v. Franklin Cty. Bd. of Educ., 766 F.2d 917, 929 (6th Cir. 1985). In these cases, most often, statistics are “coupled with anecdotal evidence of the … intent to treat the protected class unequally.” Mozee v. Am. Commercial Marine Serv. Co., 940 F.2d 1036, 1051 (7th Cir. 1991). Statistical evidence can sometimes serve by itself to establish a prima facie case in the pattern or practice context, in lieu of comparative evidence pertaining to each class member. Teamsters, 431 U.S. at 336; Hazelwood, 433 U.S. at 307–08 (“Where gross statistical disparities can be shown, they alone may in a proper case constitute prima facie proof of a pattern or practice of discrimination.") As one court explained, “strong statistics may prove a case on their own, while shaky statistics may be insufficient unless accompanied by additional evidence.” EEOC v. O & G Spring & Wire Forms Specialty Co., 38 F.3d 872, 876 (7th Cir. 1994) (citing Teamsters, 431 U.S. at 340).
While there is no “rigid mathematical formula” for determining whether a disparity is significant, Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 994–95 (1988), courts have adopted various tests to aid them in making this determination. For example, some courts have looked to whether the disparity is statistically significant. Hazelwood, 433 U.S. at 308 n.14 (an inference of discrimination will generally arise where “‘the difference between the expected value and the observed number is greater than two or three standard deviations’”) (quoting Castaneda, 430 U.S. at 496 n.17).  Other courts have looked at whether the disparity is both statistically and practically significant. See Thomas v. Metroflight, Inc., 814 F.2d 1506, 1510 n.4 (10th Cir. 1987) (suggesting that courts may require, in addition to statistical significance, that the observed disparity be substantial). Still other courts have recognized the usefulness of multiple regression analyses, a statistical tool for understanding the relationship between two or more variables where there are several possible explanations for a given outcome, which, in turn, aids in isolating the most relevant variable and determining its effect on the outcome. See, e.g., Bazemore v. Friday, 478 U.S. 385, 400 (1986) (observing the usefulness of multiple regression analysis, even one that did not include all measurable variables).
Here are a few cases in which systemic discrimination was proved:
- Latino motorists were deprived of constitutional rights as a result of being detained by a law enforcement agency conducting “saturation patrols” or “sweeps” targeting Latinos suspected of being illegally present in the country. Law enforcement deputies engaged in a pattern of racially profiling Latinos for vehicle stops. Melendres, 695 F.3d at 998 (addressing Title VI and equal protection claims).
- The deliberate and systematic exclusion of women from food server positions based on sexual stereotypes associating a “fine-dining ambience” with all–male food service may amount to a pattern or practice. While the court ultimately remanded the case because of conflicting witness testimony and conclusions drawn by the lower court, the decision set forth certain guideposts regarding the kind of evidence that may prove helpful to establish that discrimination was the defendant’s “standard operating procedure.” For example, the court noted the testimony of several witnesses who described the defendant’s active discouragement of women applying for employment. The court explained that a plaintiff may establish systemic discrimination “‘through a combination of strong statistical evidence of disparate impact coupled with anecdotal evidence of the employer’s intent to treat the protected class unequally.’ [Further,] direct evidence of an intent to discriminate’ may be used to establish a pattern or practice claim.” Joe’s Stone Crab Inc., 220 F.3d at 1285, 1287 (Title VII case) (citing Mozee, 940 F.2d at 1051, and Lujan, 766 F.2d at 929 n.15).
- Defendant’s motion for partial summary judgment was denied where the EEOC argued that the defendant’s “standard operating procedure—its regular rather than unusual practice”—was to ignore most (if not all) of its female employees’ complaints that they were individually, or as a group, being subjected to a sexually hostile and abusive environment. The alleged offensive conduct included unwelcome sexual advances, demands for sexual favors, and other offensive verbal and physical conduct of a sexual nature. The court held that the employer was aware of this possible sexual harassment and its failure to act indicated that it tolerated individual acts of sexual harassment. EEOC v. Mitsubishi Motor Mfg. of Am., 990 F. Supp. 1059, 1069 (C.D. Ill. 1998) (Title VII case).
2. Permissible Use of Race
It is critical for agencies to be aware that the exercise of a race-based motive does not mean that the recipient’s actions automatically violate Title VI. The Supreme Court has held that strict judicial scrutiny applies to a governmental entity’s intentional use of race, a standard that applies through Title VI to any recipient of Title VI funds. The Court has also held that strict scrutiny does not automatically invalidate the use of race; race may be used when the government has a compelling interest supporting its use, and that use is narrowly tailored to support the stated compelling interest. Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 720 (2007).
Moreover, agency Title VI implementing regulations recognize circumstances under which recipients’ consideration of race may be permissible. First, when “administering a program regarding which the recipient has previously discriminated against persons on the ground of race, color, or national origin, the recipient must take affirmative action to overcome the effects of prior discrimination.” 28 C.F.R. 42.104(b)(6)(1) (DOJ regulations). Second, “[e]ven in the absence of such prior discrimination, a recipient in administering a program may take affirmative action to overcome the effects of conditions which resulted in limiting participation by persons of a particular race, color, or national origin.” 28 C.F.R. 42.104(b)(6)(2) (DOJ regulations). Compelling governmental interests, thus far, have included remedying the effects of past discrimination, United States v. Paradise, 480 U.S. 149, 161 (1987), and achieving the benefits of diversity in higher education, Grutter v. Bollinger, 539 U.S. 303, 333 (2003), and law enforcement, Wittmer v. Peters, 87 F.3d 916, 920 (7th Cir. 1996). In addition, a recipient has more latitude to pursue one of these goals through actions that do not award benefits based solely on an individual’s race, color, or national origin. See Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701 (2007) (distinguishing between race conscious mechanisms to achieve diversity in public schools, such as strategic site selection of new schools, and approaches that treat specific individuals differently based on race); see also Doe ex rel. Doe v. Lower Merion Sch. Dist., 665 F.3d 524, 545–46 (3d Cir. 2011) (facially race neutral plan that involved assignment of students based on where they live did not trigger strict scrutiny).
Classifications of individuals based on race, color, or national origin cannot avoid strict scrutiny merely because the recipient asserts a very important interest, such as a public safety justification. “The gravity of the threat alone cannot be dispositive of questions concerning what means law enforcement officers may employ to pursue a given purpose.” City of Indianapolis v. Edmond, 531 U.S. 32, 42 (2000). “No matter how tempting it might be to do otherwise, [courts] must apply the same rigorous standards even where national security is at stake.” Hassan v. City of New York, 804 F.3d. 277, 306 (3d Cir. 2015). In Hassan, the Third Circuit reversed the lower court, ruling that plaintiffs had alleged a viable claim of intentional discrimination where the New York Police Department followed a facially discriminatory policy in surveilling Muslim individuals and businesses in New York and New Jersey, and that this can amount to “direct evidence of intent.” Id. at 295; see also Johnson v. California, 543 U.S. 499, 505–06 (2005) (racial classifications for penological purposes, such as controlling gang activity in prison, subject to strict scrutiny); United States v. Brignoni-Ponce, 422 U.S. 873, 885–87 (1975) (law enforcement need “does not justify stopping all Mexican-Americans to ask if they are aliens”).
Once a compelling interest is established, a recipient must still demonstrate that it has satisfied narrow tailoring; in other words, that it is using race in the most limited manner that will still allow it to accomplish its compelling interest. Parents Involved, 551 U.S. at 720. “Even in the limited circumstance when drawing racial distinctions is permissible to further [an important or] compelling state interest, [the recipient] is still ‘constrained in how it may pursue that end.’” Grutter, 539 U.S. at 333 (quoting Shaw v. Hunt, 517 U.S. 899, 908 (1996)). Strict scrutiny requires that the decision-maker “ultimately be satisfied that no workable race-neutral alternatives would” further the compelling interest “‘about as well and at tolerable administrative expense.’” Fisher v. Univ. of Tex., 133 S. Ct. 2411, 2420 (2013) (quoting Wygant v. Jackson Bd. of Ed., 476 U.S. 267, 280 n.6 (1986)). In addition, the relationship between the stated justification and the discriminatory classification must “be substantiated by objective evidence.” Patrolmen’s Benevolent Ass’n of New York v. City of New York, 310 F.3d 43, 53 (2d Cir. 2002). “[M]ere speculation or conjecture is insufficient,” id., as are appeals to “‘common sense’ which might be inflected by stereotypes,” Reynolds v. City of Chicago, 296 F.3d 524, 526 (7th Cir. 2002).
By way of illustration, in some instances police departments have used race or national origin to direct law enforcement activities, and have attempted to justify their conduct by noting that specific individuals from that race or national origin group engaged in illegal activity. Courts consistently reject this kind of stereotyping when examining expressly discriminatory law enforcement policies. See, e.g., Whren v. United States, 517 U.S. 806, 813 (“the Constitution prohibits selective enforcement of the law based on considerations such as race”). One court, in ruling that a police department’s policy of focusing on Hispanic persons in immigration enforcement was discriminatory, held “there is no legitimate basis for considering a person’s race in forming a belief that he or she is more likely to engage in a criminal violation and the requisite ‘exact connection between justification and classification’ … is lacking.” Melendres, 989 F. Supp. 2d at 901 (quoting Gratz v. Bollinger, 539 U.S. 244, 270 (2003)); see also Floyd v. City of New York, 959 F. Supp. 2d 540, 587 (S.D.N.Y. 2013) (rejecting the City’s suggestion that law-abiding members of some racial groups have a greater tendency to appear suspicious than members of other racial groups, ruling that a “stop and frisk” program was racially discriminatory).
Similarly, in Hassan, an Equal Protection Clause case involving an express religious classification, the Third Circuit held that the NYPD's blanket monitoring of the Muslim community after the September 11 attacks failed strict scrutiny because the surveillance program was not narrowly tailored. The Third Circuit compared the City's public safety justification to the infamous Korematsu case, in which the Supreme Court uncritically accepted the government's national security justification for overt discrimination, leading to the wartime imprisionment of American citizens of Japanese ancestry based solely on national origin. The Hassan court stated:
We have learned from experience that it is often where the asserted interest appears most compelling that we must be most vigilant in protecting constitutional rights. “[H]istory teaches that grave threats to liberty often come in times of urgency, when constitutional rights seem too extravagant to endure.” Skinner v. Ry. Labor Execs.’ Ass’n, 489 U.S. 602, 635 (1989) (Marshall, J., dissenting); see also Grutter, 539 U.S. at 351 (Scalia, J., concurring in part and dissenting in part) (“The lesson of Korematsu is that national security constitutes a ‘pressing public necessity,’ though the government’s use of [a suspect classification] to advance that objective must be [appropriately] tailored.”); Skinner, 489 U.S. at 635 (Marshall, J. dissenting) (“The World War II relocation- camp cases and the Red scare and McCarthy-era internal subversion cases are only the most extreme reminders that when we allow fundamental freedoms to be sacrificed in the name of real or perceived exigency, we invariably come to regret it.” (citations omitted)).
Hassan, 804 F.3d at 306–07.
Obviously, when to determine that a recipient’s consideration of race is permissible is complex, and is not extensively discussed here. Guidance documents from the Departments of Justice and Education review applicable legal principles and set out detailed considerations for educational institutions. See Dep’t of Educ. and Dep’t of Justice, “Dear Colleague” Letter on the U.S. Supreme Court ruling in Schuette v. Coalition to Defend Affirmative Action (May 6, 2014); Dep’t of Educ. and Dep’t of Justice, “Dear Colleague” Letter and Guidance Documents on the Voluntary use of Race (Dec. 2, 2011). These also may be useful in understanding how and when recipients may consider race in other contexts. Federal investigating agencies are encouraged to review applicable guidance documents and case law, and to consult their legal counsel or the Civil Rights Division for assistance applying applicable legal principles to specific situations. The Department of Education's Office for Civil Rights is also available to provide assistance about the use of race in the educational context.
3. Intentional Discrimination by a Third Party
Hostile environment harassment is another form of intentional discrimination prohibited by Title VI not discussed here extensively. When the recipient does not create the hostile environment, but a third party, who neither speaks for nor represents the recipient, is responsible, the hostile environment framework focuses on the recipient’s obligation to respond adequately to the third party’s discriminatory conduct. Both courts and federal agencies have addressed this circumstance in the context of hostile environment discrimination in schools.
A recipient violates Title VI if (1) a third party (e.g., a fellow student) harasses a program participant or beneficiary based on race, color, or national origin and the harassing conduct is sufficiently serious to deny or limit the individual’s ability to participate in or benefit from the program or activity (i.e., the harassment creates a hostile environment); (2) the recipient knew or reasonably should have known about the alleged harassment, i.e., actual or constructive notice; and (3) the recipient fails to take prompt and effective steps reasonably calculated to end the harassment, eliminate the hostile environment, prevent its recurrence, and address its effects, as appropriate. A recipient is liable under Title VI for its own conduct when it fails to take adequate steps to address discriminatory harassment.
Liability in private suits for monetary damages involving student-on-student harassment lies “only where the funding recipient acts with deliberate indifference to known acts of harassment in its programs or activities.” Davis v. Monroe Cty. Sch. Bd., 526 U.S. 629, 633 (1999). Often, but not always, termed “deliberate indifference” cases, the standard of proof has been most commonly applied to harassment claims, particularly sex- and race-based claims. However, courts have recognized the standard in cases involving other forms of discriminatory conduct. See, e.g., Blunt v. Lower Merion School District, 767 F.3d 247, 271–73 (3d Cir. 2014) (plaintiffs may establish a school district’s liability under Title VI for racially motivated student assignments through a deliberate indifference theory).
Similarly, a private plaintiff or investigating agency may be able to use evidence that a recipient knew or should have known about a third party’s intentionally discriminatory conduct and failed to act despite that knowledge.
 Unlike when seeking judicial enforcement, private parties may file administrative complaints under any theory of liability, including disparate impact. Section VII of the Title VI Legal Manual provides an analysis of the disparate impact theory.
 U.S. Cons. amend. XIV, § 1.
 42 U.S.C. § 2000e et seq.
 Note that the analysis under these civil rights law are not always the same, particularly to the extent that the Equal Protection Clause affords different levels of protection to classifications based on sex and disability vs. race, color, and national origin.
 At times in this section "race" is used to refer to "Race, color, and national origin." This shorthand is used merely for ease of discussion and should not be read as a limitation on the applicability of the principles discussed.
 Vill. of Arlington Heights v. Metro. Housing Dev. Corp., 429 U.S. 252, 266-68 (1977).
 The McDonnell- Douglas framework refers to McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973)
 See, e.g., Gebser v. Lago Vista Indep. Sch. Dist., 524 U.S. 274, 286 (1998) (“[Title VI] is parallel to Title IX …. The two statutes operate in the same manner ….”); Liese v. Indian River Cty. Hosp. Dist., 701 F.3d 334, 346 (11th Cir. 2012) (“Title IX, like the [Rehabilitation Act] was modeled after Title VI, and the text of all three acts [is] virtually identical ….”); Darensburg v. Metro. Transp. Comm’n, 636 F.3d 511, 519 (9th Cir. 2011) (looking to Title VII jurisprudence to analyze Title VI claims).
 Price Waterhouse has been superseded by statute in the employment discrimination context under Title VII, but as discussed below, its framework remains instructive when considering how to prove mixed motives cases in other civil rights contexts.
 See Michigan Elliott-Larsen Civil Rights Act, MCL 37.2101 et seq. (2016; Jackson v. Quanex Corp., 191 F.3d 647 (6th Cir. 1999)(When an employer is liable under the Michigan Civil Rights Act, it would also be liable under Title VII).
 Though the Arlington Heights test was developed to detect discriminatory intent in the context of a Fourteenth Amendment Equal Protection claim, the test also applies to claims of intentional discrimination under some federal statutes, including Title VI. See Pac. Shores Props., 730 F.3d at 1158 n.21; see also Gallagher v. Magner, 619 F.3d 823, 833 (8th Cir. 2010)(Fair Housing Act case applying the Arlington Heights factors); Hallmark Developers, Inc. v. Fulton Cty., 466 F.3d 1276, 1283-84 (11th Cir. 2006)(same); Tsombanidis v. W. Haven Fire Dep't, 352 F.3d 565, 579-80 (2d Cir. 2003)(same in Fair Housing Act and Americans with Disabilities Act contexts).
 For a detailed case analysis of statistical evidence, circumstantial evidence, the strength of each, and the cumulative picture of intent presented by both types of evidence together in the Title VII context, see Gay, 694 F.2d at 555-56.
 Similary, an agency may be able to use impact evidence under the deliberate indifference framework, originally developed to analyze hostile environment harassment claims, to show that the recipient knew a federally protected right was substantially likely to be violated and failed to act despite that knowledge. This approach is closely related to the Arlington Heights framework. As in the cases discussed in this section, foreseeability or knowledge of harm is a key feature of this method of proof. See infra section C.3.
The Pryor court partially distinguished Feeney, 442 U.S. at 256, in which the Court refused to find that a Massachusetts veterans' preference statute deprived women of equal protection of the laws. It noted that the NCAA had actual notice and knowledge of the impact on the minority students, while the Court in Feeney could no infer that the "legislature almost certainly was aware" that the law benefiting veterans would disadvantage women. Pryor, 288 F.3d at 564.
 In a subsequent proceeding, the court granted summary judgment for the defendants on the issue of intentional discrimination under Title VI by noting that “assuming, arguendo, that Plaintiffs are correct that ‘[t]he disparate impact of [issuing the permit to the defendant] was clearly [foreseeable]’ to [the defendants], Pls.’ Opp. at 71, such a foreseeable impact is of no aid to Plaintiffs at this juncture because it, alone, is insufficient to establish a constitutional violation.” S. Camden Citizens in Action v. N.J. Dep’t of Envtl. Prot., No. Civ. A. 01-702 (FLW), 2006 WL 1097498 at *36 (D.N.J. Mar. 31, 2006) (citing Penick, 443 U.S. at 465). In so ruling, the court found insufficient evidence of Arlington Heights factors alleged at the motion to dismiss stage, such as a history of discrimination on the part of the defendant. S. Camden, 2006 WL 1097498 at *26–28. The court determined that, in the absence of the other Arlington Heights factors raised at the motion to dismiss stage, foreseeable impact alone is insufficient to demonstrate intent. Penick has cautioned that “disparate impact and foreseeable consequences, without more, do not establish a constitutional violation.” Penick, 443 U.S. at 464. See also Dayton Bd. of Educ. v. Brinkman, 443 U.S. 526, 536 n.9 (1979) (foreseeable adverse impact may be relevant evidence in proving purposeful discrimination, but foreseeability by itself has not been held to make out a case of purposeful discrimination).
The elements of a prima facia case are the same under both Title VI and VII. Paul v. Theda Med. Cty., Inc., 465 F.3d 790, 794 (7th Cir. 2006); Fuller v. Rayburn, 161 F.3d 516, 518 (8th Cir. 1998).
 The DOJ regulations quoted here are similar to those of other agencies.
 Dep't of Justice and Dep/'t of Educ., "Dear Collegue" Letter on the Nondiscriminatory Administration of School Discipline (Jan. 8, 2014), available at http://www2.ed.gov/about/offices/list/ocr/letters/colleague-201401-title-vi.html.
 The report of investigation is located on the following website: http://www.justice.gov/crt/special-litigation-section-cases-and-matters (search "antelope"; last visited Sept. 15, 2016).
 Note that "the absence of statistical evidence [will not] invariably prove fatal in every pattern or practice case. [In employment cases,] [w]here the overall nunber of employees in small, anecdotal evidence may suffice." In re W. Dist. Xerox Litig., 850 F. Supp. 1079, 1084 (W.D.N.Y. 1994); accord, Pitre v. Western Elec. Co., 843 F.2d 1262, 1268 (10th Cir. 1988); Haskell v. Kaman Corp., 743 F.2d. 113, 119 (2d Ci. 1984). Conversely, in certain cases "a plaintiff's statistucal evidence alone might constitute a prima facia case." Coates v. Johnson & Johnson, 756 F.2d. 524, 532 n.6 (7th Cir. 1985) (citing Segar v. Smith, 738 F.2d 1249, 1278 (D.C. Cir. 1984)). "Neither statistical nor anecdotal evidence is automatically entitled to reverence to the exclusion of the other." Id. at 533. However, "[w]hen one type of evidence is missing altogether, the other must be correspondingly stronger for plaintiffs to meet their burden." In re W. Dist. Xerox Litig., 850 F. Supp at 1085. Compare Chisholm v. USPS, 665 F.2d 482, 495 (4th Cir. 1981) (twenty class plaintiffs was sufficient to support the statistical evidence) with Ste. Marie v. E. R.R. Ass'n, 650 F.2d 395, 406 (2d Cir. 1981) (seven discriminatory acts coupled with problematic statistical evidence were insufficient to support finding pattern or practice discrimination).
 However, "[t]here is no minimum statistical threshold" mandating that plaintiff has demonstrated a violation. Waisome v. Port Auth. of N.Y. & N.J., 948 F.2d 1370, 2376 (2d Cir. 1991); accord Chin v. Port Auth. of N.Y. & N.J., 685 F.3d 135, 153 (2d Cir. 2012). Courts should take a "'case-by-case approach' in judging the significance or substantiality of disparities, one that considers not only statistics but also all the surrounding facts and circumstances." Waisome, 948 F.3d at 1376; Chin, 685 F.3d at 13 (quoting Waisome).
 Korematsu v. United States, 324 U.S. 885 (1944).
 Dep't of Educ. Off. for Civ. Rts., "Dear Colleague" Letter: Harrassment and Bullying, (Oct. 26, 2010), available at http://www2.ed.gov/about/offices/list/ocr/letters/colleague-201010.pdf; see also Dep't of Educ. Complaint Resolution Letter, Richmond Heights School District (OH), No. 15-11-1134 (May 11, 2012); Revised Sexual Harassment Guidance: Harassment of Students by School Employees, Other Students, or Third Parties, 66 Fed. Reg. 5512-01 (Jan. 19, 2001).